Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
49 C H A P T E R 7 7.1 Research Approach 7.1.1 Basic Concepts A row of small buildings such as the detached houses in Fig- ure 33 acts as a series of small noise barriers with gaps in between them, which reduces sound levels at receivers behind the row. In FHWA TNM 2.5, one could choose to model each house as a barrier object. With some exceptions, this approach is generally not used except for model comparison to field measurements. Even then, when each house is modeled as a barrier, the TNM algorithms do not account for the reflection of sound from the highway off the sides of the buildings to receivers behind them. The FHWA TNM building row object simulates a row of houses as a single long barrier with a low uniform transmis- sion loss, something akin to a porous noise barrier. TNM dif- fracts a portion of the sound energy over the top of the row and allows a portion to pass uniformly through the row. The specific locations of the gaps between the buildings are not defined and thus the effect of their exact location in the row is not computed. Unlike FHWA TNM Barriers, building rows cannot be perturbed up and down from the average height. The building row may be defined as series of connected straight line segments, defined by x, y, and z coordinates of the segment endpoints. The base of the building row defines the terrain over which the sound passes from source to receiver and thus affects the computation of the interactions with the ground in the propagation calculations. Because a building row defines the location of the ground and because it causes sound to be diffracted over it, its location relative to the roadway and receiver is also important, as are the elevations of the ground at the roadway and receiver, and the height above ground of the receiver. The closer the building row is to the source or receiver, all else remaining the same, the greater the diffraction attenuation will be. Other needed input parameters are the following: (1) aver- age height of the buildings above the user-specified ground elevations; and (2) building percentage, which is the percent- age of the line defined by the row that is blocked by the row (or 100% minus the percentage of gaps [gap fraction] between the houses). The allowable range is 20% to 80%. TNM computes sound-level reductions based on user- defined location, length, average height, and percentage of area blocked. More than one building row may be modeled in a run. Where there is more than one row of buildings between the highway and the receivers, a simplified calculation is used. As described in the âTraffic Noise Model: Fre- quently Asked Questions FAQsâ section on the FHWA noise web site:29 TNM first identifies all building rows that interrupt the effective source-receiver path. Rows that do not interrupt the propagation path are ignored. For each row that interrupts the path, TNM determines which building row has the most effective attenu- ation at the 630 Hz frequency band. For this building row, the actual attenuation is calculated for all 1â3-octave frequency bands. For each remaining row that interrupts the propagation path, an attenuation of 1.5 dB is applied to each 1â3-octave band. The maximum attenuation for any number of building rows has been set to 10 dB. For a listing of maximum attenuation for each 1â3-octave band please refer to Table 13 on Page 100 in the TNM Tech Manual. Table 13 from the FHWA TNM Technical Manual30 is presented in Appendix F, which is available on the NCHRP Project 25-34 web page at http://apps.trb.org/cmsfeed/TRB NetProjectDisplay.asp?ProjectID=2986. As with many of the other FHWA TNM objects, there are always concerns about over-modeling of building rows that does not improve accuracy. As noted above, detailed 1â3-octave band attenuations are only calculated for the most effective Building Rows 29 Traffic Noise Model: Frequently Asked Questions FAQs, FHWA website: www. fhwa.dot.gov/environment/noise/traffic_noise_model/tnm_faqs. 30 FHWA Traffic Noise Model, Technical Manual, FHWA No. FHWA-PD-96-010, February 1998.
50 building row, with other building rows simply adding 1.5 dB attenuation per 1â3 octave band. Nonetheless, the FHWA TNM building row object is a very important parameter in noise impact assessment, affecting how quickly sound levels drop off with distance from the road. The building row object is also, possibly, an important parameter in noise barrier reasonable- ness assessment, potentially affecting the number of residences benefited by a noise barrier because of the interplay between building row attenuation and noise barrier attenuation. Unfortunately, the noise reduction calculated by FHWA TNM for building rows cannot be easily field validated because building rows do not spatially locate the real-world gaps through which sound passes. In addition, as one gets deeper into a community, background noise and refractive meteoro- logical effects caused by wind shear, wind direction, and tem- perature lapse rateânone of which are modeled in FHWA TNM 2.5âbecome much more important in determining the overall sound level measured at a site, making model validation difficult and often impossible. The goals of this research were to test and refine current FHWA guidelines on modeling building rows, specifically as related to â¢ Height. â¢ Building percentage. â¢ Use of an FHWA TNM building row or individual building barrier objects. â¢ Effect of receiver location behind a row of houses. Two other items were considered: â¢ How does the presence of a building row affect the calcu- lated noise reduction from a noise barrier along the edge of the road? â¢ How much noise reduction is computed when the building rows are perpendicular to the modeled roadway? 7.1.2 Research Tasks There are two basic candidate modeling techniques that were evaluated: â¢ Use of the FHWA TNM building row object for rows paral- lel to the highway and rows perpendicular to the highway through sensitivity testing and use of the available valida- tion data sets. â¢ Use of the FHWA TNM barrier object for modeling rows of houses or other small building as individual building barriers. The work began with a survey of practitioners, which iden- tified one very well-documented field validation study of modeling detached houses as individual barriers, conducted by and for the Maryland State Highway Administration. Four other field validation studies previously conducted by the NCHRP Project 25-34 research team were also identified. These studies all had very good sets of validation dataâ measured sound levels with concurrent traffic classification counts and speed measurements. The FHWA TNM runs were also made available to the research team. The Maryland State Highway Administration project was different from the others in that it used the approach of modeling each house as an FHWA TNM barrier object. Following a review of the surveyed material, a sensitivity analysis was conducted for generic situations to examine the variation in the modeled wayside sound-level results due to variations in the different TNM building row input param- eters, specifically the percentage of a row that is blocked by the buildings, the average height of the buildings, and the number of rows that are modeled. Situations were studied both with and without barriers, and with rows that are both parallel and perpendicular to the roadway. The next step was to analyze the available TNM valida- tion data sets. For two older studies, the FHWA TNM 1.0b runs were converted to FHWA TNM 2.5 format. For the other studies, the original modeling had been done with FHWA TNM 2.5. Separate model runs were made using building rows and building barriers, and, in two cases, for specific pavement types. The modeled levels were com- pared to the measurement results, keeping in mind that any differences in measured and modeled results could not be solely attributed to the use of the FHWA TNM building row object. Further, the TNM building row object is, by its nature, an approximation because it does not account for the actual locations of the gaps between the buildings. Some of the runs were modified to study the calculated effects of varying the receiver positions behind the building row. The building row modeling guidelines were refined based on the research results. Figure 33. Detached houses in front of a highway on an embankment.31 31 Source: Bowlby & Associates, Inc.
51 7.2 Outcome of the ResearchâBest Practices and How to Implement Them for a Noise Study or TNM 7.2.1 Distance from Building Rows and Percentage of Blockage The results showed the following: â¢ As the distance from the building row (and the roadway) increases, the amount of noise reduction decreases. â¢ As the building percentage increases, the amount of noise reduction increases. â¢ As a result of the above two findings, simple guidance may not be sufficient. 7.2.2 Sensitivity to Building Row Height Table 9 shows the change in noise reduction (not the actual noise reduction) behind a single building row as a function of building height for different building percentages. The table may be used to decide how precise the modeling of the height needs to be for a given situation. The shaded values are for differences of half a dB or more. The values should be viewed as indicators of the effect, not as absolute noise reduction differences. The effects will be specific to the situation being modeled. Also, Table 9 is for a receiver in the middle of the building row. Near the ends of the row, the sound coming from the unshielded area beyond the end of the row will reduce the effect of changes in the building row parameters. The results show the following: â¢ For 20% to 40% blockage, a change in height of 5 ft causes little change in the noise reduction regardless of how far back the receiver is behind the building row. â¢ For higher building percentages, the change in noise reduc- tion is dependent on the distance behind the building row. The maximum difference for a 5-ft height changeâfrom 20 to 25 ft and from 25 to 30 ftâis less than 2 dB in these modeled cases. Sensitivity runs were completed for one building row par- allel to an eight-lane road (four lanes in each direction) with shoulders and a 30-ft-wide grassy median. Each lane was mod- eled with mixed traffic traveling at 60 mph. The building row was 70 ft from the edge of the near travel lane. Three build- ing heights were chosen (20, 25, and 30 ft) to represent one- story and two-story buildings with an intermediate height. The percentage of blockage was varied from 20% to 80% in 10% increments, along with a 0% case (no row). Receivers were modeled up to 1,000 ft from the road. Full details are in Appendix F. Figure 34 shows the noise reductions as a function of building percentage for the 20-ft height, and Figure 35 shows results for the 30-ft height. In both cases, up close to the building row, the noise reduction varies from about 1 dB for 20% blockage to over 6 dB for 80% blockage. Up close, the path of sound through the building row is more dominant than the diffracted sound over the top of the row, hence the greater sensitiv- ity to building percentage. As the distance back from the building row increases, the amount of diffraction attenu- ation over the top of the building row decreases, such that the path over the top of the row begins to dominate the total received level: â¢ For the 20-ft-high building row, by just over 330 ft behind the row, the noise reduction is 1 dB or less regardless of the building percentage. â¢ In contrast, for the 30-ft-high building row, there is a more gradual decrease in the noise reduction as the dis- tance away from the building row increases. For example, at 330 ft behind the row, the noise reduction for 80% blockage is 3.7 dB, roughly 2.7 dB greater than that for a 20-ft-high building row. 7.2.3 Sensitivity to Building Percentage 22.214.171.124 One Building Row The building row noise reduction is also a function of the building percentage. Table 10 compares the change in the noise reduction from one building percentage to a value that is 10% greater (e.g., from 20% to 30%) for single building row heights of 20, 25, and 30 ft. Table 10 may be used to assess how precise the modeling of the building percentage needs to be for a given situation. The values given in Table 10 are not the noise reductions, but the changes from one case to the next. Shading iden- tifies differences of 0.5 dB or more. The values provided in Table 10 should be viewed as indicators of the size of the effect, not as absolute noise reduction differences. The effects will be specific to the situation being modeled. Also, Table 10 is for a receiver in the middle of a building row. Near the ends of the row, the sound coming from the unshielded area beyond the end of the row will reduce the effect of changes in the building percentage. The results show that the greater the building row height, the greater the difference in going from one build- ing percentage to another percentage that is 10% higher or lower. For building percentages of 20 to 60%, a variation of Â±10% around a given percentage will produce a noise reduction difference of less than 1 dB at all of the studied distances
52 Distance to Building Row, ft Noise Reduction Differences between Different Building Percentages, dB 20% 30% 40% 50% 60% 70% 80% Building Row Height Change from 20 ft to 25 ft 10 0.0 0.0 â0.1 â0.1 â0.1 â0.2 â0.3 30 â0.1 0.0 â0.2 â0.2 â0.2 â0.3 â0.5 50 â0.1 â0.1 â0.2 â0.3 â0.5 â0.7 â0.9 70 â0.2 â0.2 â0.2 â0.4 â0.5 â0.7 â1.0 90 â0.1 â0.2 â0.3 â0.5 â0.6 â0.8 â1.2 110 â0.2 â0.3 â0.3 â0.5 â0.6 â0.9 â1.2 130 â0.2 â0.2 â0.4 â0.5 â0.8 â1.0 â1.4 170 â0.2 â0.3 â0.5 â0.7 â1.0 â1.4 â1.7 210 â0.3 â0.4 â0.6 â0.8 â1.1 â1.4 â1.8 250 â0.3 â0.4 â0.6 â0.8 â1.1 â1.4 â1.8 290 â0.2 â0.4 â0.6 â0.8 â1.1 â1.3 â1.7 330 â0.2 â0.4 â0.6 â0.8 â1.0 â1.3 â1.6 430 â0.3 â0.3 â0.5 â0.7 -0.9 â1.1 â1.3 530 â0.2 â0.3 â0.5 â0.6 â0.8 â1.0 â1.3 730 â0.2 â0.4 â0.5 â0.7 â0.9 â1.0 â1.3 930 â0.2 â0.4 â0.5 â0.6 â0.8 â1.0 â1.2 Building Row Height Change from 25 ft to 30 ft 10 0.0 â0.1 0.0 0.0 0.0 0.0 â0.1 30 0.0 â0.1 0.0 0.0 â0.1 â0.2 â0.2 50 0.0 â0.1 â0.1 â0.1 â0.1 â0.2 â0.3 70 0.0 0.0 â0.1 â0.1 â0.2 â0.3 â0.5 90 0.0 0.0 â0.1 â0.1 â0.2 â0.3 â0.5 110 0.0 0.0 â0.1 â0.2 â0.3 â0.4 â0.7 130 0.0 â0.1 â0.2 â0.3 â0.3 â0.6 â0.8 170 â0.1 â0.2 â0.2 â0.3 â0.4 â0.6 â0.9 210 â0.1 â0.1 â0.2 â0.3 â0.5 â0.6 â0.9 250 â0.1 â0.1 â0.2 â0.4 â0.5 â0.7 â1.0 290 â0.2 â0.2 â0.2 â0.4 â0.5 â0.8 â1.1 330 â0.1 â0.2 â0.3 â0.4 â0.6 â0.8 â1.1 430 â0.1 â0.3 â0.4 â0.5 â0.7 â1.0 â1.3 530 â0.2 â0.3 â0.4 â0.6 â0.8 â1.0 â1.3 730 â0.2 â0.2 â0.4 â0.5 â0.7 â1.0 â1.3 930 â0.1 â0.2 â0.3 â0.5 â0.7 â0.9 â1.2 Building Row Height Change from 20 ft to 30 ft 10 0.0 â0.1 â0.1 â0.1 â0.1 â0.2 â0.4 30 â0.1 â0.1 â0.2 â0.2 â0.3 â0.5 â0.7 50 â0.1 â0.2 â0.3 â0.4 â0.6 â0.9 â1.2 70 â0.2 â0.2 â0.3 â0.5 â0.7 â1.0 â1.5 90 â0.1 â0.2 â0.4 â0.6 â0.8 â1.1 â1.7 110 â0.2 â0.3 â0.4 â0.7 â0.9 â1.3 â1.9 130 â0.2 â0.3 â0.6 â0.8 â1.1 â1.6 â2.2 170 â0.3 â0.5 â0.7 â1.0 â1.4 â2.0 â2.6 210 â0.4 â0.5 â0.8 â1.1 â1.6 â2.0 â2.7 250 â0.4 â0.5 â0.8 â1.2 â1.6 â2.1 â2.8 290 â0.4 â0.6 â0.8 â1.2 â1.6 â2.1 â2.8 330 â0.3 â0.6 â0.9 â1.2 â1.6 â2.1 â2.7 430 â0.4 â0.6 â0.9 â1.2 â1.6 â2.1 â2.6 530 â0.4 â0.6 â0.9 â1.2 â1.6 â2.0 â2.6 730 â0.4 â0.6 â0.9 â1.2 â1.6 â2.0 â2.6 930 â0.3 â0.6 â0.8 â1.1 â1.5 â1.9 â2.4 Table 9. Change in noise reduction behind a single building row as a function of building height for different building percentages.
53 -1 0 1 2 3 4 5 6 7 0 100 200 300 400 500 600 700 800 900 1000 Di ffe re nc e i n L A eq 1h fr om 0 % ca se , d B Distance from edge of near travel lane, 20% blockage 30% blockage 40% blockage 50% blockage 60% blockage 70% blockage 80% blockage Figure 34. Noise reduction as a function of building percentage for a single 20-ft-high building row 70 ft from edge of an eight-lane roadway. -1 0 1 2 3 4 5 6 7 0 100 200 300 400 500 600 700 800 900 1000 Di ffe re nc e in L A eq 1h fr om 0 % c as e, d B Distance from edge of near travel lane, 20% blockage 30% blockage 40% blockage 50% blockage 60% blockage 70% blockage 80% blockage Figure 35. Noise reduction as a function of building percentage for a single 30-ft-high building row 70 ft from edge of an eight-lane roadway.
54 Distance Behind Building Row, ft Noise Reduction Differences between Different Building Percentages, dB From 20 to 30% From 30 to 40% From 40 to 50% From 50 to 60% From 60 to 70% From 70 to 80% Building Row Height of 20 ft 10 0.5 0.6 0.8 0.9 1.1 1.5 30 0.6 0.5 0.7 0.8 1.0 1.3 50 0.5 0.5 0.6 0.7 0.8 1.1 70 0.5 0.5 0.5 0.7 0.7 0.9 90 0.4 0.4 0.5 0.6 0.7 0.7 110 0.4 0.5 0.4 0.6 0.5 0.7 130 0.4 0.3 0.5 0.4 0.5 0.5 170 0.3 0.3 0.3 0.3 0.3 0.4 210 0.3 0.2 0.3 0.2 0.3 0.2 250 0.3 0.2 0.2 0.2 0.2 0.1 290 0.2 0.2 0.2 0.1 0.2 0.0 330 0.2 0.1 0.2 0.1 0.1 0.1 430 0.2 0.1 0.1 0.1 0.0 0.0 530 0.2 0.1 0.1 0.0 0.0 â0.1 730 0.1 0.1 0.0 0.0 0.0 â0.2 930 0.0 0.1 0.0 â0.1 â0.1 â0.3 Building Row Height of 25 ft 10 0.5 0.7 0.8 0.9 1.2 1.6 30 0.5 0.7 0.7 0.8 1.1 1.5 50 0.5 0.6 0.7 0.9 1.0 1.3 70 0.5 0.5 0.7 0.8 0.9 1.2 90 0.5 0.5 0.7 0.7 0.9 1.1 110 0.5 0.5 0.6 0.7 0.8 1.0 130 0.4 0.5 0.6 0.7 0.7 0.9 170 0.4 0.5 0.5 0.6 0.7 0.7 210 0.4 0.4 0.5 0.5 0.6 0.6 250 0.4 0.4 0.4 0.5 0.5 0.5 290 0.4 0.4 0.4 0.4 0.4 0.4 330 0.4 0.3 0.4 0.3 0.4 0.4 430 0.2 0.3 0.3 0.3 0.2 0.2 530 0.3 0.3 0.2 0.2 0.2 0.2 730 0.3 0.2 0.2 0.2 0.1 0.1 930 0.2 0.2 0.1 0.1 0.1 â0.1 Building Row Height of 30 ft 10 0.6 0.6 0.8 0.9 1.2 1.7 30 0.6 0.6 0.7 0.9 1.2 1.5 50 0.6 0.6 0.7 0.9 1.1 1.4 70 0.5 0.6 0.7 0.9 1.0 1.4 90 0.5 0.6 0.7 0.8 1.0 1.3 110 0.5 0.6 0.7 0.8 0.9 1.3 130 0.5 0.6 0.7 0.7 1.0 1.1 170 0.5 0.5 0.6 0.7 0.9 1.0 210 0.4 0.5 0.6 0.7 0.7 0.9 250 0.4 0.5 0.6 0.6 0.7 0.8 290 0.4 0.4 0.6 0.5 0.7 0.7 330 0.5 0.4 0.5 0.5 0.6 0.7 430 0.4 0.4 0.4 0.5 0.5 0.5 530 0.4 0.4 0.4 0.4 0.4 0.5 730 0.3 0.4 0.3 0.4 0.4 0.4 930 0.3 0.3 0.3 0.3 0.3 0.2 Table 10. Changes in noise reduction for 10% changes in building percentages for a single building row.
55 behind the building row. For a difference of less than 0.5 dB, the needed distance behind the building row is a function of building row height: â¢ 20 ft highâbeyond 130 ft behind the building row. â¢ 25 ft highâbeyond 250 ft behind the building row. â¢ 30 ft highâbeyond 430 ft behind the building row. For building percentages of 60 to 80% and for building row heights of 20 to 30 ft, the accuracy of the estimated building percentage is a bit more important than at the lower building percentages. Differences in noise reduction for a variation of Â± 10% are as large as 1.5 dB close behind the building row, but drop down to under 1 dB at distances from 70 to 210 ft as the building row height increases from 20 to 30 ft. Another way to use the results is to determine the error in simply using a building percentage of 50% regardless of the actual percentage. Figure 36 shows the differences in noise reduction for various building percentages compared to 50% blockage for a 20-ft-high building row: â¢ The error is less than 1 dB for 40% and 60% blockage at all distances behind the building row. â¢ For 30% and 70%, the receiver has to be about 170 ft or farther behind the building row for the error to be under 1 dB. â¢ At about 330 ft back and beyond, the 20% and 80% cases could be represented as 50% with an error of under 1 dB. Appendix F includes the results for the 25- and 30-ft-high building rows. The patterns are similar to the 20-ft case, except that the differences are somewhat greater, especially as one moves farther behind the row. 126.96.36.199 Two Building Rows A second building row should be modeled if a second row is present in the study area. Modeling a second building row will have important effects in reducing the predicted level at a receiver compared to modeling only a single row. Care should be used in approximating the height and building percentage of both rows because FHWA TNM will determine which row is more effective and use it as the primary basis for its noise reduction calculations. That choice can vary by receiver- roadway pair. Cases were studied for a single 12-ft-wide roadway, with the first building row 70 ft from the edge of the travel lane and the second building row 150 ft back from the first, a typical distance in going across a residential street with houses on relatively small lots. Figure 37 shows the predicted noise reductions for both rows at a height of 20 ft for building percentages varying from 20 to 80%. Notice how the noise reduction decreases with increasing distance from the first building row, steps back up for receivers behind the second row, and then decays again with increasing distance behind the second row. -4 -3 -2 -1 0 1 2 3 4 0 100 200 300 400 500 600 700 800 900 1000 Di ffe re nc e in L Ae q1 h fr om c as e to c as e, d B Distance from edge of near travel lane, From 50 to 20% From 50 to 30% From 50 to 40% From 50 to 60% From 50 to 70% From 50 to 80% Figure 36. Differences in noise reduction for various building percentages compared to 50% blockage for a single 20-ft-high building row 70 ft from the edge of an eight-lane roadway.
56 For two building rows, the noise reductions are sensitive to building row height as the distance back from the second row increases. Close in behind the second row, the differences in noise reduction are generally small for the different heights except for the high building percentages. Of interest is the amount of the increase in the noise reduc- tion when going behind the second rowânearly 3.5 dB for the 80% blockage case. The FHWA FAQ states that when two rows are present, the program computes the attenuation for each building row on a 1â3-octave band basis, chooses the more effective one, and then adds simply 1.5 dB to the A-weighted level for the less effective one. These results illustrate that when the receiver is moved behind the second row, the program is choosing the second row as more effective than the first row because the receiver is directly behind the second row with an associated very large diffraction angle. The change in level is not as simple as subtracting 1.5 dB from the level computed for the first building row. Building percentage also affects the amount of noise reduc- tion and the step-up in that noise reduction when going behind the second row. Both values decrease as building per- centage decreases. For example, the step-up in noise reduction for the 20% blockage case is only 1.8 dB compared to 3.5 dB for 80% blockage. Beyond 400 ft, the total noise reduction for the two-row case drops below 2 dB, which is still 1 to 1.5 dB greater than the single row case at these distances. Figure 38 presents the differences in noise reduction between building percentages that differ by 10% for the two- row case with the 20-ft building row height. A 10% change in the first-row blockage results in a change in the noise reduc- tion from 0.4 dB for the low building percentages to 1.5 dB for the highest building percentages. The changes behind the second row (which is located at 220 ft from the edge of the roadway) range from 0.4 to 0.9 dB right behind the row, but quickly drop to under a 0.5 dB at greater distances back. Appendix F provides the results for the 25- and 30-ft-high building rows. For these greater heights, the decay in the noise reduction is not as rapid as for the 20-ft height, both in front of and behind the second row. 188.8.131.52 Three Building Rows Modeling a third building row will have important effects in reducing the predicted sound level at a receiver compared to modeling only a single row or two rows and should be done if a third row is present in the study area. Care should be used in approximating the height and building percentage of each row because FHWA TNM will determine the most effec- tive row and use it as the primary basis for its noise reduction calculations. That choice can vary by receiver-roadway pair. For the tests for the three-row case, the third building row was located at a distance of 150 ft back from the second row in the previous tests, or 370 ft from the edge of the travel lane. Figure 39 shows the predicted noise reductions (differ- ences from the 0% blockage case) for the three-row scenario. This figure is for 20-ft building row heights for building per- centages varying from 20 to 80%. The pattern described for -1 0 1 2 3 4 5 6 7 0 100 200 300 400 500 600 700 800 900 1000 Di ffe re nc e in L A eq 1h fr om 0 % c as e, d B Distance from edge of near travel lane, 20% blockage 30% blockage 40% blockage 50% blockage 60% blockage 70% blockage 80% blockage Figure 37. Noise reduction as a function of building percentage for two 20-ft-high building rows near a single 12-ft-wide roadway.
57 -1 0 1 2 3 4 5 6 7 0 100 200 300 400 500 600 700 800 900 1000 Di ffe re nc e in L A eq 1h fr om 0 % c as e, d B Distance from edge of near travel lane, 20% blockage 30% blockage 40% blockage 50% blockage 60% blockage 70% blockage 80% blockage Figure 39. Noise reduction as a function of building percentage for three 20-ft-high building rows near a single 12-ft-wide roadway. -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 0 100 200 300 400 500 600 700 800 900 1000 Di ffe re nc e in L Ae q1 h be tw ee n ca se s, d B Distance from edge of near travel lane, From 20 to 30% From 30 to 40% From 40 to 50% From 50 to 60% From 60 to 70% From 70 to 80% Figure 38. Differences in noise reduction for 10% incremental increases of building percentage for two 20-ft-high building rows near a single 12-ft-wide roadway.
58 the two-row case is present: the noise reduction decreasing with increasing distance from the first building row. The noise reduction then steps back up for receivers behind the second row, and then decays again with increasing distance from the second row until it steps back up again for receivers behind the third row. As with the two-row case, there is a large increase in the noise reduction when going behind the second rowânearly 3.5 dB for the 80% blockage case. There is also a 4 dB increase when going behind the third row. As the building percentage decreases, the noise reduction decreases, and the stepping up when going behind the second and third rows decreases as well. Beyond 500 ft, the total noise reduction for the three-row case varies between 2 and 3.5 dB across all of the building percentages. Appendix F includes the results for the 25- and 30-ft-high building rows. For the greater heights, the decay in the noise reduction is not as rapid as it is for the 20-ft height, both in front of and behind the second row. 7.2.4 Building Row Effect on Noise Barrier Noise Reduction The predicted sound level behind a noise barrier decreases in the presence of an intervening building row, in addition to the âno-barrierâ level decreasing. The noise reduction pro- vided by that barrier may increase or decrease slightly over the case without a building row, depending on the barrier height and the building row parameters. A 1.5-dB difference was modeled in the noise reduction for a 20-ft-high noise barrier between the 30% and 70% blockage cases for a 30-ft- high building row. The change in the barrier noise reduction could change the barrier acoustical design and decisions on barrier feasibility and reasonableness depending on whether or not receptors behind the building row(s) were impacted and/or benefited in accordance with the criteria in a state highway agencyâs traffic noise policy. The eight-lane road scenario was modified to include a noise barrier just off the shoulder. Barrier heights of 12 to 28 ft were tested in addition to a âno-barrierâ case. A single building row was located 70 ft from the edge of the near travel lane. For these tests, building row heights of 20 and 30 ft and building percentages of 30, 50 and 70% were used. Figure 40 shows the predicted sound levels for one of the cases: the 20-ft-high noise barrier and the 30-ft-high build- ing row. Note that both the no-barrier and with-barrier levels decrease as the building percentage increases. In the no-barrier cases, the with-building-row levels are roughly 4 to 5 dB lower than the no-building-row case. In the with- barrier cases, the with-building-row levels are 2 to 3 dB lower than the no-building-row levels. Figure 41 is the companion graph of the differences in the no-barrier and with-barrier levelsâthe noise reduction provided by the 20-ft barrier. The no-building-row noise reductions are very similar to the 50% blockage case for this particular example. 45 50 55 60 65 70 75 80 0 100 200 300 400 500 600 700 800 900 1000 L A eq 1h , d BA Distance from edge of near travel lane, No barrier, 0% No barrier, 30% No barrier, 50% No barrier, 70% 20-, 0% 20-, 30% 20-, 50% 20-, 70% Figure 40. Sound level with and without a 20-ft-high noise barrier along the edge of the shoulder of an eight-lane roadway with a 30-ft-high building row located 70 ft from the edge of the near travel lane, for various building percentages.
59 The other cases are in Appendix F and show similar effects on both the with-barrier and no-barrier levels and the resul- tant overall noise reductions. 7.2.5 Building Rows Perpendicular to the Roadway Rows of houses can also be perpendicular to the high- wayâalong streets that end at the highway right-of-way, in a cul-de-sac, or intersecting with a collector road that is parallel and adjacent to the highway. The noise reductions for perpendicular building rows are not as large as reductions for building rows parallel to the highway. Exposure to noise coming from beyond the ends of the rows and coming through the gaps for local streets or adjoining back yards appears to dominate the modeled levels. Another option to modeling the scenario where rows of houses are perpendicular to the road is to model FHWA TNM building rows parallel to the highway, cutting across the local streets and front and back yards instead of running along them. The building percentage for such building rows parallel to the highway would most likely be substantially lower than for the perpendicular rows in order to accommodate the depth of yards reaching from the local streets to the rear property lines. Figure 42 shows a case in a TNM plan view of four building rows representing rows of houses on both sides of two perpen- dicular streets. In this scenario, the building rows are modeled 150 ft apart on either side of the streets, and the distance across the backyards between the two âinteriorâ rows is 200 ft. Front yard receivers are located 30 ft in front of the building row, and backyard receivers are located 50 ft behind the building row. The modeling of both front yard and backyard receiv- ers for the same houses would not typically be done, but was used in this analysis to test the differences. The receivers are spaced assuming lot widths of 100 ft. Thus, a building percent- age of 50% would mean a 50-ft-wide house with 25 ft of yard on either side. Building percentages of 30, 50, and 70% were tested for building row heights of 20 ft and 30 ft. For the outermost backyard receivers exposed to upstream or downstream noise the greatest noise reduction was only 1.3 dB for 20-ft-high building rows and 1.7 dB for 30-ft-high building rows, both for 70% blockage. For the two sets of internal front yard receivers, the greatest noise reduction was 2.2 dB for the 20-ft-high building rows and 3.3 dB for the 30-ft-high building rows, both for 70% blockage. 7.2.6 Modeling as Individual Building Barriers Instead of as Building Rows Some noise analysts prefer to model each house as an individual noise barrier rather than using the FHWA TNM building row object. As a standard procedure, the Maryland State Highway Administration models rows of houses as individual noise barriers, which it calls âbuilding-barriers.â Figure 43 shows a portion of a noise study area for one Maryland State Highway Administration project.32 Figure 44 5 6 7 8 9 10 11 12 13 14 15 16 17 0 100 200 300 400 500 600 700 800 900 1000 N oi se re du c on w ith b ar rie r, dB Distance from edge of near travel lane, 20-, 0% 20-, 30% 20-, 50% 20-, 70% Figure 41. Noise reduction for a 20-ft-high noise barrier along the edge of the shoulder of an eight-lane roadway with a 30-ft-high building row located 70 ft from the edge of the near travel lane, for various building percentages. 32 Rummel, Klepper & Kahl, LLP, and Maryland State Highway Administration Noise Abatement Design & Analysis Team, âHO317A21 US 29 Widening, Type I, Technical Noise Analysis Report,â 2010.
60 Figure 42. FHWA TNM plan view of four perpendicular building rows adjacent to an eight-lane highway. Figure 43. Portion of Noise Study Area G for US 29 study.33 33 Schematic in Figure 43 is from Rummel, Klepper & Kahl, LLP, and Maryland State Highway Administration Noise Abatement Design & Analysis Team, âHO317A21 US 29 Widening, Type I, Technical Noise Analysis Report,â 2010.
61 shows the corresponding Maryland State Highway Administra- tion modeling of the houses as building barriers on the right and the comparative modeling of the houses as FHWA TNM building rows on the left. In the case shown in Figure 44, the analyst chose to model three sides of each house as barrier segments. Choice of sides is usually dictated by which sides appear to offer the most noise reduction to receivers behind them. For a receiver close to the ârearâ side, that side may be most effective. For a build- ing close to the roadway, the side facing the roadway may be the most effective. One option is to model all four sides and let FHWA TNM compute the combined effect using its double diffraction algorithms. However, if more than two âhighest path pointsâ are detected by FHWA TNM for any given receiver-to- roadway path (including noise abatement barriers and ter- rain), only the two most effective points will be considered for that path. In addition, since FHWA TNM does not compute the effects of sound reflections off the sides of the buildings, detailed modeling of all four sides of a row of buildings will not necessarily mean a more accurate sound-level prediction behind the modeled buildings. Further, the accuracy of the modeling is not improved by modeling every small turn in the faÃ§ade of a building; straight line segments for each faÃ§ade are sufficient. Modeling as barriers does take more time than modeling as building rows. Useful guidance on how to model houses as building barriers, working in Bentleyâs InRoadsÂ© roadway design computer-aided design (CAD) program, was provided by an in-house Maryland State Highway Administration consultant,34 summarized as follows: Houses are represented by either three-sided barriers or two- sided barriers, considering which orientation would likely pro- vide the best shielding. The house shape is outlined and then âliftedâ above the ground elevations on the surface. Sometimes, roof elevations (if available) are used to help define the height. Other times, an estimated height is used to create either an irregular top or a level one. The primary interest is in modeling the most massive part of the structure (below the roof line) to be conservative, as opposed to modeling a sloped roof. Finally, the preference is to place receptors in the gap spaces between the houses for greatest exposure to the modeled traffic noise. If modeling only two sides of a building, the modeler needs to choose the two sides carefully so as not to expose the receiver to a greater view of the roadway (and thus, more noise) than would occur in the real world. Figure 44. TNM plan view with building rows (left) and building barriers (right), Noise Study Area G of US 29 widening. 34 Based on email from Matthew G. Mann, Sr., P.E., in-house consultant from The Wilson T. Ballard Company on the Maryland State Highway Administration Noise Abatement Design & Analysis Team.
62 184.108.40.206 Case Study Comparisons of Measured and Modeled Data For this analysis, five different projects with a total of eight noise study areas were examined for a comparison between modeling detached houses using building rows and mod- eling using building barriers. Sound-level measurement results with concurrent traffic counts were available. The counts were used in modeling the sites to allow comparison to the measured levels. Building height and percentage of blockage were site specific. There were a total of 34 receiv- ers in the models that were at least partially shielded from roadway noise by intervening houses. For one project, three repetitions of the measurements were made. Each receiver had predictions run with the âAverageâ pavement type in FHWA TNM. For two of the projects with portland cement concrete (PCC) pavement, predictions were also run with PCC pavement and normalized to a reference microphoneâs measured levels. From the five cases, 138 discrete comparisons were made between measured levels and (1) levels from a model using building rows and (2) levels from a model using building barriers. A detailed presentation on the projects and the measurement and modeling results is provided in Appen- dix F. In general, the overall results and comparisons seemed reasonable for three of the projects. However, for two of the projects, generally poor agreement was found between the measured levels and levels predicted by both modeling methods. With the broad range of site conditions and results, it is difficult to generalize. However the following observations are made: â¢ Modeling each house as a barrier generally gave lower predicted sound levels than modeling a building row, ranging from an increase of less than 1 dB to a 5 dB decrease. The average reduction in sound level across all cases was 1.5 dB. â¢ In over 80% of the 138 comparisons of measured and modeled levels using the building barrier approach, the modeled level was within 3 dB of the measured level. In contrast, roughly 67% of the modeled building row com- parisons were within 3 dB of the measured levels. â¢ For the non-normalized results, neither of the two mod- eling methods was substantially better than the other in terms of producing FHWA TNM results that were closer to the measured sound levels. â¢ In the cases where normalizing to a reference microphone was done, the building barrier approach provided better agreement with the measurements than the building row approach. This last finding supports the idea of modeling rows of houses as building barriers, but not to the point of being recommended instead of using building rows. Model- ing using barriers should be considered if the modeler is attempting to validate an FHWA TNM to measured lev- els. However, the need for validation can be questioned for receivers that have one or more rows of houses between them and the road. One instance in which such validation might be more important would be where a state highway administrationâs traffic noise policy includes criteria for determining the feasibility and reasonableness of a noise barrier that are based on a percentage of all of a study areaâs impacted receptors and benefited receptors, as compared to only the first-row receptors. Even then, the reliability of that validation can be questioned, given the greater distance from the road for such receivers and the effects of meteoro- logical conditions and other sources of noise on the mea- sured levels. What is apparent from the results is that if a state high- way administration decides to model shielding by rows of houses with building barriers instead of building rows, it should do so consistently on all projects and should have carefully defined procedures for the modeling in terms of the individual barriers, placement of the receiver points, and possibly the terrain elevation between the barri- ers. The latter two factors are described in the following sections. 220.127.116.11 Effect of Receiver Location Behind Building Barriers Receiver placement can be an important factor for the predicted sound level when using building barriers, but it also may not be critical, depending in part on how close the receivers are placed to the houses, the size of the houses, and the amount of blockage provided by the houses. Differences as small as 0.3 dB and as large as 4 dB were seen in shifting the receiver location laterally in the tested cases. One concern is that the predicted sound level can be sensitive to a receiver location near the gap between two building barriers or directly behind one of the barriers. In particular, variations in the modeling techniqueâbuilding rows versus building barriers, distance back from the houses, height of the houses, and receiver location relative to the gaps between the housesâcan change findings in an analysis. A typical state highway administration policy value for defining noise impact for a proposed project is 66 dBA. The levels in the tests fell on either side of that criterion, meaning that the receptors represented by these FHWA TNM receivers would be found as impacted in one analysis and not impacted in another.
63 This effect was tested using four FHWA TNM runs based on three of the modeled sites from the analysis of modeling houses as building rows versus modeling them as building barriers. Figure 45 shows two of the tested cases, with one set of receivers (labeled 1 through 7) farther from the intervening row of houses and the second (labeled A through G) closer to the row. In this case, the analyst chose to model the walls of the houses facing away from the road (to the left in the figure) as the more effective facades because the houses were closer to the receivers than to the roadway. The results varied. For the two cases shown in Figure 45, there is only a 0.3 dB variation in levels for the more distant receivers, but a 4 dB range for the closer receivers. The highest sound level is in front of the gap between the houses and the lowest sound level directly in front of one of the houses. For the other two tested cases, the levels varied by only 0.4 dB in one case and 1.1 dB in the other. For the cases shown in Figure 45, the building percent- age was 70% and the houses were relatively long (deep into the lot) relative to their width. Farther back, it was much less likely that a receiver would be placed where there was not blockage of the line of sight through the gaps. Closer in, it was more likely that there would be a direct line of sight of the roadway through a gap, causing a larger variation in the predicted levels than farther back. 18.104.22.168 Effect of Terrain Elevation in the Gaps between Building Barriers As a final note, in certain situations care needs to be taken to properly model the ground elevation in between the individual building barriers. With a building row, the ground elevation is defined along the entire row based on the user-input elevations of the segment points for the building row. However, in defin- ing individual barriers, FHWA TNM is only provided with the ground elevations of the barrier segment points, not the ground elevation in the gaps between the individual barriers. Figure 46 shows two âskew sectionâ views of paths from a receiver to the highway, one that passes through the building and captures the elevation of the base of the building and one that passes through the gap and misses that ground eleva- tion information. The result could be incorrect calculations of the amount of terrain shielding or excess ground attenua- tion for sound paths in the gaps and, thus, an incorrect total predicted sound level at a receiver behind the buildings. In those cases where the terrain does vary, it might be necessary to model short FHWA TNM terrain line objects connecting between each building barrier. Alternatively, a separate terrain line could be defined slightly in front of or behind the building barriers that properly models the ground elevation of the adjacent building barriersâ ground points. Figure 45. I-65 south with two sets of receivers at different distances behind the row of buildings.
64 Figure 46. FHWA TNM plan view (top) and skew section views through a building barrier (middle) and through a gap (bottom) where the ground elevation has not been defined.